[1/2] treewide: Add a User Agent for CURL requests

This will help identifying robot usage across PW instances.

Signed-off-by: Aaron Conole <aconole@redhat.com>
---
 ci_mon             |  2 +-
 github_get_logs.sh |  8 ++++----
 github_mon         |  2 +-
 github_restart     |  6 ++++--
 jenkins_lib.sh     |  5 +++--
 post_pw.sh         |  8 ++++----
 pw_mon             | 14 +++++++-------
 7 files changed, 24 insertions(+), 21 deletions(-)
  

On 1/22/24 18:26, Aaron Conole wrote:
> This will help identifying robot usage across PW instances.

Thanks, Aaron!

I didn't test, but the patch seems fine to me in general.

But what do you think about adding a common string to all the
requests, so it's easier to identify them?  With this change
the user agent is different for different parts of the system.
Maybe do something like 'User-Agent: xxx (pw-ci)', so all the
requests have '(pw-ci)' part in common?

Best regards, Ilya Maximets.
  

On Mon, Jan 22, 2024 at 12:26 PM Aaron Conole <aconole@redhat.com> wrote:
>
> This will help identifying robot usage across PW instances.
Thanks Aaron for the quick patch!

My only suggestion is to prepend '0daybot' to the user agent (or
anything close that you like). That way we know the requests are
coming from us. It might be easier for Jeremy to identify than just
"pw-post" alone for example

Again thanks for the quick turn around!


Acked-by: Michael Santana <msantana@redhat.com>

>
> Signed-off-by: Aaron Conole <aconole@redhat.com>
> ---
>  ci_mon             |  2 +-
>  github_get_logs.sh |  8 ++++----
>  github_mon         |  2 +-
>  github_restart     |  6 ++++--
>  jenkins_lib.sh     |  5 +++--
>  post_pw.sh         |  8 ++++----
>  pw_mon             | 14 +++++++-------
>  7 files changed, 24 insertions(+), 21 deletions(-)
>
> diff --git a/ci_mon b/ci_mon
> index 2e30da4..31de0ab 100755
> --- a/ci_mon
> +++ b/ci_mon
> @@ -159,7 +159,7 @@ for I in travis github obs dummy; do
>          author=$(echo $SERIES_LINE | cut -d\| -f4)
>          email=$(echo $SERIES_LINE | cut -d\| -f5)
>
> -        PATCHDATA=$(curl -s $url)
> +        PATCHDATA=$(curl -A "ci-mon-${PROJECT}" -s $url)
>
>          patch_id="$(get_patch_id_by_series_id_and_sha "$series_id" "$shasum" "$pw_chk_instance")"
>          if [ "X$patch_id" == "X" ]; then
> diff --git a/github_get_logs.sh b/github_get_logs.sh
> index 9f44840..19c2278 100755
> --- a/github_get_logs.sh
> +++ b/github_get_logs.sh
> @@ -74,14 +74,14 @@ print_errored_logs_for_commit () {
>      run_id="$(echo "$run_meta" | jq -r ".id")"
>
>       # Get the real logs url, download logs, and unzip logs
> -    logs_url="$(curl -s -S -H "${AUTH}" -H "${APP}" "${redirect_url}" -I | \
> +    logs_url="$(curl -A "github-logs-crawler" -s -S -H "${AUTH}" -H "${APP}" "${redirect_url}" -I | \
>          grep -i 'Location: '| sed 's/Location: //i' | tr -d '\r')"
> -    curl -s -S "$logs_url" -o "build_logs_series_${series_id}.zip"
> +    curl -A "github-logs-crawler" -s -S "$logs_url" -o "build_logs_series_${series_id}.zip"
>      unzip -o -q "build_logs_series_${series_id}.zip" -d "build_logs_series_${series_id}"
>
>      # Get the names of the failed jobs and the steps that failed
>      tmp_url="$GITHUB_API/repos/$repo_name/actions/runs/${run_id}/jobs"
> -    jobs_results="$(curl -s -S -H "${AUTH}" -H "${APP}" "${tmp_url}")"
> +    jobs_results="$(curl -A "github-logs-crawler" -s -S -H "${AUTH}" -H "${APP}" "${tmp_url}")"
>      jobs_results="$(echo "$jobs_results" | jq "[.jobs[] | \
>          select(.conclusion==\"failure\") | {name, failed_step: .steps[] | \
>          select(.conclusion==\"failure\") | {name, conclusion, number}}]")"
> @@ -135,7 +135,7 @@ repo_name=$(echo "$repo_name" | sed -e 's@%2F@/@g' -e 's,%40,@,g')
>
>  # Get GHA runs
>  tmp_url="$GITHUB_API/repos/$repo_name/actions/runs?per_page=9000"
> -all_runs="$(curl -s -S -H "${AUTH}" -H "${APP}" "${tmp_url}")"
> +all_runs="$(curl -A "github-logs-crawler" -s -S -H "${AUTH}" -H "${APP}" "${tmp_url}")"
>  runs=$(echo $all_runs | jq -rc ".workflow_runs[] | select(.head_branch == \"series_$series_id\")")
>  not_found="$(echo "$runs" | jq -rc ".message")"
>  if [ "$not_found" == "Not Found" ]; then
> diff --git a/github_mon b/github_mon
> index 8670ce8..311c7bd 100755
> --- a/github_mon
> +++ b/github_mon
> @@ -117,7 +117,7 @@ while IFS="|" read -r series_id patch_id patch_url patch_name sha patchwork_inst
>          tmp_url="$GITHUB_API/repos/$repo_name/actions/runs?per_page=9000"
>          if [ "$tmp_url" != "$prev_url" ]; then
>              prev_url="$tmp_url"
> -            all_runs="$(curl -s -S -H "${AUTH}" -H "${APP}" "${tmp_url}")"
> +            all_runs="$(curl -A "github-mon-${pw_project}" -s -S -H "${AUTH}" -H "${APP}" "${tmp_url}")"
>          fi
>
>          runs=$(echo "$all_runs" | jq -rc ".workflow_runs[] | select(.head_branch == \"series_$series_id\")")
> diff --git a/github_restart b/github_restart
> index d5d63b9..18af978 100755
> --- a/github_restart
> +++ b/github_restart
> @@ -114,7 +114,8 @@ if [ "X$runid" == "X" ]; then
>      fi
>
>      comma=""
> -    for job in $(curl -s -S -H "${AUTH}" -H "${APP}" \
> +    for job in $(curl -A "github-restart-${pw_project}" -s -S -H "${AUTH}" \
> +                      -H "${APP}" \
>                        "https://api.github.com/repos/${reponame}/actions/runs?head_sha=${sha}" | \
>                       jq -rc ".workflow_runs[] ${workflow_select} .id")
>      do
> @@ -126,7 +127,8 @@ fi
>  echo -n "{\"results\":["
>  comma=""
>  for job in $(echo "$runid" | sed 's/,/ /'); do
> -    result=$(curl -s -X POST -L -S -H "${AUTH}" -H "${APP}" \
> +    result=$(curl -A "github-restart-${pw_project}" -s -X POST -L -S \
> +                  -H "${AUTH}" -H "${APP}" \
>                    "https://api.github.com/repos/${reponame}/actions/runs/$job/rerun")
>      msg=$(echo "$result" | jq -rc '.message')
>
> diff --git a/jenkins_lib.sh b/jenkins_lib.sh
> index fc7cb1d..09c176c 100644
> --- a/jenkins_lib.sh
> +++ b/jenkins_lib.sh
> @@ -41,12 +41,12 @@ if [ "X$jenkins_token" != "X" ]; then
>      jenkins_credentials="$jenkins_user:$jenkins_token"
>  fi
>
> -jenkins_crumb_value=$(curl -s "http://${jenkins_credentials}@${jenkins_url}/crumbIssuer/api/xml?xpath=concat(//crumbRequestField,\":\",//crumb)")
> +jenkins_crumb_value=$(curl -A "jenkins-sub-${PROJECT}" -s "http://${jenkins_credentials}@${jenkins_url}/crumbIssuer/api/xml?xpath=concat(//crumbRequestField,\":\",//crumb)")
>
>  function jenkins_check_for_job() {
>      local jenkins_job=jenkins_${pw_project}_job
>
> -    curl -s -f -X GET \
> +    curl -A "jenkins-sub-${PROJECT}" -s -f -X GET \
>           "http://${jenkins_credentials}@${jenkins_url}/job/${!jenkins_job}/config.xml" >/dev/null
>  }
>
> @@ -86,6 +86,7 @@ function jenkins_submit_series() {
>          }'
>
>      curl -s -f -X POST \
> +         -A "jenkins-sub-${PROJECT}" \
>           -H "${jenkins_crumb_value}" \
>           --data token="${!jenkins_job_token}" \
>           --data-urlencode json="$json_data" \
> diff --git a/post_pw.sh b/post_pw.sh
> index e98e8ab..fe2f41c 100755
> --- a/post_pw.sh
> +++ b/post_pw.sh
> @@ -53,7 +53,7 @@ send_post() {
>          return 0
>      fi
>
> -    report="$(curl -sSf "${link}")"
> +    report="$(curl -A "pw-post" -sSf "${link}")"
>      if [ $? -ne 0 ]; then
>          echo "Failed to get proper server response on link ${link}" 1>&2
>          return 0
> @@ -79,7 +79,7 @@ send_post() {
>      fi
>
>      # Get reports from patch
> -    checks="$(curl -sSf -X GET \
> +    checks="$(curl -A "pw-post" -sSf -X GET \
>          --header "Content-Type: application/json" \
>          "$api_url")"
>      if [ $? -ne 0 ]; then
> @@ -105,7 +105,7 @@ send_post() {
>          \"description\": \"$description\"\
>      }"
>
> -    curl -sSf -X POST \
> +    curl -A "pw-post" -sSf -X POST \
>          -H "Authorization: Token ${token}" \
>          --header "Content-Type: application/json" \
>          --data "$data" \
> @@ -120,7 +120,7 @@ send_post() {
>  }
>
>  year_month="$(date +"%Y-%B")"
> -reports="$(curl -sSf "${mail_archive}${year_month}/thread.html" | \
> +reports="$(curl -A "pw-post" -sSf "${mail_archive}${year_month}/thread.html" | \
>           grep -i 'HREF=' | sed -e 's@[0-9]*<LI><A HREF="@\|@' -e 's@">@\|@')"
>  if [ $? -ne 0 ]; then
>      echo "Failed to get proper server response on link ${reports}" 1>&2
> diff --git a/pw_mon b/pw_mon
> index ee77256..7ff57cf 100755
> --- a/pw_mon
> +++ b/pw_mon
> @@ -146,7 +146,7 @@ function check_new_series() {
>
>      GET_URL="http://${INSTANCE}/api/series/?project=${PROJECT}&since=${SINCE}"
>
> -    response=$(curl -s "$userpw" "$GET_URL")
> +    response=$(curl -A "pw-mon-${PROJECT}" -s "$userpw" "$GET_URL")
>
>      series_info=$(echo "$response" | jq -rc '.[] | (.id|tostring) + ";" + .url + ";" + .submitter.name + ";" + .submitter.email + ";" + (.received_all|tostring)')
>
> @@ -162,7 +162,7 @@ function check_new_series() {
>  function check_completed_series() {
>      get_uncompleted_jobs_as_line "$pw_instance" "$pw_project" | while IFS=\| read -r id url submitter_name submitter_email; do
>          echo "Checking on series: $id"
> -        local RESPONSE=$(curl -s "$userpw" "$url" | jq -rc '.received_all')
> +        local RESPONSE=$(curl -A "pw-mon-${PROJECT}" -s "$userpw" "$url" | jq -rc '.received_all')
>          if [ "$RESPONSE" = "true" ]; then
>              echo "Setting series $id to completed"
>              series_id_set_complete "$pw_instance" "$id"
> @@ -183,8 +183,8 @@ function check_superseded_series() {
>      local pw_instance="$1"
>      series_get_active_branches "$pw_instance" | while IFS=\| read -r series_id project url repo branchname; do
>          # first query the patch states
> -        local last_patch_url=$(curl -s "$userpw" "$url" | jq -rc '.patches[-1].url')
> -        local patch_state=$(curl -s "$userpw" "$last_patch_url" | jq -rc '.state')
> +        local last_patch_url=$(curl -A "pw-mon-${PROJECT}" -s "$userpw" "$url" | jq -rc '.patches[-1].url')
> +        local patch_state=$(curl -A "pw-mon-${PROJECT}" -s "$userpw" "$last_patch_url" | jq -rc '.state')
>
>          # now check to see if the patch should even be reported...
>          if [ "$patch_state" = "superseded" -o "$patch_state" = "rejected" -o "$patch_state" = "accepted" \
> @@ -214,7 +214,7 @@ function check_patch_for_retest_request() {
>      local pw_project="$2"
>      local patch_url="$3"
>
> -    local patch_json=$(curl -s "$userpw" "$patch_url")
> +    local patch_json=$(curl -A "pw-mon-${PROJECT}" -s "$userpw" "$patch_url")
>      local patch_comments_url=$(echo "$patch_json" | jq -rc '.comments')
>      local patch_id=$(echo "$patch_json" | jq -rc '.id')
>      local series_id=$(echo "$patch_json" | jq -rc '.series[].id')
> @@ -231,7 +231,7 @@ function check_patch_for_retest_request() {
>      fi
>
>      if [ "Xnull" != "X$patch_comments_url" ]; then
> -        local comments_json=$(curl -s "$userpw" "$patch_comments_url")
> +        local comments_json=$(curl -A "pw-mon-${PROJECT}" -s "$userpw" "$patch_comments_url")
>          local seq_end=$(echo "$comments_json" | jq -rc 'length')
>          if [ "$seq_end" -a $seq_end -gt 0 ]; then
>              seq_end=$((seq_end-1))
> @@ -250,7 +250,7 @@ function check_series_needs_retest() {
>      local pw_instance="$1"
>      local pw_project="$2"
>
> -    local series_list=$(curl -s "$userpw" "http://${pw_instance}/api/series/?project=${pw_project}&state=new&state=rfc&state=under-review&archived=false&order=-id")
> +    local series_list=$(curl -A "pw-mon-${PROJECT}" -s "$userpw" "http://${pw_instance}/api/series/?project=${pw_project}&state=new&state=rfc&state=under-review&archived=false&order=-id")
>      local n=$(echo "$series_list" | jq -rc 'length')
>
>      if [ "Xnull" == "X$n" -o "X" == "X$n" ]; then
> --
> 2.41.0
>
  

Ilya Maximets <i.maximets@ovn.org> writes:

> On 1/22/24 18:26, Aaron Conole wrote:
>> This will help identifying robot usage across PW instances.
>
> Thanks, Aaron!
>
> I didn't test, but the patch seems fine to me in general.
>
> But what do you think about adding a common string to all the
> requests, so it's easier to identify them?  With this change
> the user agent is different for different parts of the system.
> Maybe do something like 'User-Agent: xxx (pw-ci)', so all the
> requests have '(pw-ci)' part in common?

Makes sense.  Done.

> Best regards, Ilya Maximets.